llama-3.1-8B-grpo / all_results.json
AlistairPullen's picture
Model save
3db6010 verified
{
"total_flos": 0.0,
"train_loss": 0.08981622397834212,
"train_runtime": 6983.3902,
"train_samples": 45000,
"train_samples_per_second": 3.093,
"train_steps_per_second": 0.064
}